Agents - KFUPM Open Courseware
advertisement
ICS-381 Principles of Artificial Intelligence Weeks 3-4 Intelligent Agents Chapter 2 Dr.Tarek Helmy El-Basuny Dr. Tarek Helmy, ICS-KFUPM 1 Review: What is AI? Views of AI fall into four categories: Thinking humanly Thinking rationally Acting humanly Acting rationally We will focus on "acting rationally" Acting rationally: rational agent Rational behavior: doing the right thing The right thing: which is expected to maximize goal achievement, given the available information. Dr. Tarek Helmy, ICS-KFUPM 2 Topics to be Covered What is Agent? Agents and Environments Rational Agents PEAS (Performance measure, Environment, Actuators, Sensors) Examples: Agents and Task Environments Environment Types Agent Types  Table Driven Agents use a percept-action table in memory to find the next action.  Reflex agents respond immediately to percepts.  Goal-based agents act in order to achieve their goals.  Utility-based agents maximize their own utility function.  Learning agents improve their performance over time. Dr. Tarek Helmy, ICS-KFUPM 3 Agents and Environments An agent is anything that can be viewed as perceiving its environment through sensors and acting upon that environment through actuators. Sensors Environment Percepts ? Agent Actions Effectors Agent’s percepts sequence: the complete history of every thing the agent has ever perceived. The agent function maps from percept histories to actions: [f: P* Æ A] The agent program runs on the physical architecture to produce f Agent = Architecture + Program Dr. Tarek Helmy, ICS-KFUPM 4 What do you mean? Sensors/Percepts and Effectors/Actions Humans ÂSensors: Eyes (vision), ears (hearing), skin (touch), tongue (gustation), nose (olfaction), etc.. ÂPercepts: At the lowest level – electrical signals from these sensors After preprocessing – objects in the visual field (location, textures, colors, …), auditory streams (pitch, loudness, direction), … ÂEffectors: legs, arms, limbs, eyes, tongue, … ÂActions: rotate left, walk, run, carry an object, … Dr. Tarek Helmy, ICS-KFUPM 5 Example: Automated Taxi Driving System Percepts: Video, speedometer, engine sensors, keyboard input, microphone, … Actions: Steer, accelerate, brake, horn, speak/display, … Goals: Maintain safety, reach destination, maximize profits (fuel, tire wear), obey laws, provide passenger comfort, … Environment: Duhran street, freeways, traffic, weather, customers, …, etc. Different aspects of driving may require different types of agent programs! Dr. Tarek Helmy, ICS-KFUPM 6 Vacuum-Cleaner World Percepts: location and state of the environment, Actions: Left, Right, Suck, NoOp There are only two locations: square A and square B, the agent perceives which square it is in and whether there is dirt in the square. It can choose to move left, right or suck. Dr. Tarek Helmy, ICS-KFUPM 7 Rational Agents Rational Agent: one that does the right thing Rational Agent: Should select an action that is expected to maximize its performance measure, based on the evidence provided by the percept sequence and whatever built-in knowledge the agent has. Performance measure: An objective criterion for success of an agent's behavior  E.g., performance measure of a vacuum-cleaner agent could be amount of dirt cleaned up, amount of time taken, amount of electricity consumed, amount of noise generated, etc. • When do we evaluate the performance is also important. If we measured how much dirt the agent cleaned up in the first hour of the day, we would be rewarding those agents that start fast and punishing those work late. Agents can perform actions in order to modify future percepts so as to obtain useful information (information gathering, exploration) An agent is autonomous if its behavior is determined by its own percepts & experience (with ability to learn and adapt) without depending solely on build-in knowledge Dr. Tarek Helmy, ICS-KFUPM 8 Rationality of Agents Depends on The performance measure that defines degree of success. Every thing the agent has seen so fare (the percept sequence). What the agent knows about the environment. The actions the agents can perform. For example, if an agent does not look both ways before crossing a busy road, then its percept sequence will not tell it that there is a truck coming at high speed. Crossing without looking is so dangerous. So an ideal agent would look before crossing to maximize the performance Dr. Tarek Helmy, ICS-KFUPM 9 Ideal Rational Agents Collaborative Collaborative Learning LearningAgents Agents Cooperate Cooperate Learn Learn Smart Smart Agents Agents Autonomous Autonomous Collaborative Collaborative Agents Agents Dr. Tarek Helmy, ICS-KFUPM Interface Interface Agents Agents 10 Task Environment Before we design a rational agent we need to specify a task environment  A problem specification for which the agent is a solution PEAS: to specify a task environment  Performance measure  Environment  Actuators  Sensors Dr. Tarek Helmy, ICS-KFUPM 11 PEAS: Taxi Driver Agent Example: Agent = Taxi driver  Performance measure: Safe, fast, legal, comfortable trip, maximize profits  Environment: Roads, other traffic, pedestrians, customers  Actuators: Steering wheel, accelerator, brake, signal, horn  Sensors: Cameras, sonar, speedometer, GPS, odometer, engine sensors, keyboard Dr. Tarek Helmy, ICS-KFUPM 12 PEAS: Medical Diagnosis Agent Example: Agent = Medical diagnosis system Performance measure: Healthy patient, minimize costs, Environment: Patient, hospital, staff Actuators: Screen display (questions, tests, diagnoses, treatments, referrals) Sensors: Keyboard (entry of symptoms, patient's answers) Dr. Tarek Helmy, ICS-KFUPM 13 PEAS: Interactive English Tutor Agent: Interactive English tutor Performance measure: Maximize student's score on test Environment: Set of students Actuators: Screen display (exercises, suggestions, corrections) Sensors: Keyboard Dr. Tarek Helmy, ICS-KFUPM 14 PEAS: Part-Picking Robot Example: Agent = Part-picking robot Performance measure: Percentage of parts in correct bins Environment: Conveyor belt with parts, bins Actuators: Jointed arm and hand Sensors: Camera, joint angle sensors Dr. Tarek Helmy, ICS-KFUPM 15 Environment Types Fully observable (vs. partially observable): An agent's sensors give it access to the complete state of the environment at each point in time. Deterministic (vs. stochastic): A deterministic environment is one in which any action has a single guaranteed effect — there is no uncertainty about the state that will result from performing an action. Episodic (vs. sequential): the performance of an agent is dependent on a number of discrete episodes, with no link between the performance of an agent in different scenarios. Static environment: A static environment is one that can be assumed to remain unchanged except by the performance of actions by the agent. A dynamic environment: is one that has other processes operating on it, and which hence changes in ways beyond the agent’s control. Discrete (vs. continuous): A limited number of distinct, clearly defined percepts and actions, (e.g. Chess game). How do we represent or abstract or model the world? Single agent (vs. multi-agent): An agent operating by itself in an environment. Dr. Tarek Helmy, ICS-KFUPM 16 Environment Types Task environment Observable Deterministic/ Episodic/ Static/ Discrete/ Stochastic Sequential Dynamic Continuous Agents Crossword puzzle fully deterministic sequential static discrete single Chess with clock fully strategic sequential semi discrete multi Poker partial stochastic sequential static discrete multi Back gammon fully stochastic sequential static discrete multi Taxi driving partial stochastic sequential dynamic continuous multi Medical diagnosis partial stochastic sequential dynamic continuous single Image analysis fully deterministic episodic semi continuous single Part-picking robot partial stochastic episodic dynamic continuous single Refinery controller partial stochastic sequential dynamic continuous single Interact. Eng. tutor partial stochastic sequential dynamic discrete multi Dr. Tarek Helmy, ICS-KFUPM 17 Some Agent Types • • Goal of AI: • Given a PEAS task environment, • Construct agent function f, • Design an agent program that implements f on a particular architecture Four basic agent types in order of increasing generality • Table-Driven Agents – Use a percept sequence/ action table in memory to find the next action. They are implemented by a (large) lookup table. • Simple Reflex agents – Are based on condition-action rules and implemented with an appropriate production (rule-based) system. They are stateless devices, which do not have memory of past world states. • Model-based reflex – Have internal state, which is used to keep track of past states of the world. • Agents with Explicit Goals – Are agents, which in addition to state information have a kind of goal information that describes desirable situations. Agents of this kind take future events into consideration. • Utility-Based Agents – Base their decision on classic axiomatic utility-theory in order to act rationally. Dr. Tarek Helmy, ICS-KFUPM 18 Table Driven Agent Current state of decision process Lookup- Table for entire history function REFLEX_VACUUM_AGENT( percept ) returns an action (location,status) = UPDATE_STATE( percept ) if status = DIRTY then return SUCK; else if location = A then return RIGHT; else if location = B then return LEFT; Dr. Tarek Helmy, ICS-KFUPM 19 Table Driven Agent • Problems – Huge number of possible percepts (consider an automated taxi with a camera as the sensor) => lookup table would be huge. – Takes long time to build the table. – Not adaptive to changes in the environment; requires entire table to be updated if changes occur. Dr. Tarek Helmy, ICS-KFUPM 20 Simple Reflex Agents • It differs from the lookup table agent in that the condition (that determines the action) is already higher-level interpretation of the percepts. • If car in front is breaking then initiate barking • A simple reflex agent works by finding a rule whose condition matches the current situation (as defined by the percept) and then doing the action associated with that rule. function SIMPLE-REFLEX-AGENT(percept) returns action static: rules, a set of condition-action rules state Å INTERPRET-INPUT (percept) rule Å RULE-MATCH (state,rules) action Å RULE-ACTION [rule] return action Dr. Tarek Helmy, ICS-KFUPM First match. No further matches sought. Only one level of deduction. 21 Simple Reflex Agents Current state of decision process NO MEMORY: Fails if environment is partially observable Example: Vacuum cleaner world Dr. Tarek Helmy, ICS-KFUPM 22 Simple Reflex Agent • Simple Reflex Agent should has condition-action pairs defining all possible condition-action rules necessary to interact in an environment • • e.g. if car-in-front-is-breaking then initiate breaking Problems – Table is still too big to generate and to store (e.g. taxi) – Takes long time to build the table – No knowledge of non-perceptual parts of the current state – Not adaptive to changes in the environment; requires entire table to be updated if changes occur – Looping: Can’t make actions conditional Dr. Tarek Helmy, ICS-KFUPM 23 Model-Based Reflex Agents New: Description of current world state Sometimes it is unclear what to do without a clear goal • In some applications, making a decision based on the current percepts only is not the right one. • The agent has to maintain some sort of internal state, these states are: • How the world evolves ( what the state of the surrounded cars), • What will happen if this action does (going left or right is ok or no). Dr. Tarek Helmy, ICS-KFUPM 24 Goal-Based Agent • Keeping track of the current state is often not enough. It needs to add goals to decide which situations are good. • Choose actions so as to achieve a given goal. • A goal is a description of a desirable situation • May have to consider long sequences of possible actions before deciding if goal is achieved, involves consideration of the future. Dr. Tarek Helmy, ICS-KFUPM 25 Goal-Based Agents • • Goals provide reason to prefer one action over the other. We need to predict the future: we need to plan & search New Dr. Tarek Helmy, ICS-KFUPM 26 Utility-based Agents • • • Some solutions to goal states are better than others. Which one is best given by a utility function. Which combination of goals is preferred? New Dr. Tarek Helmy, ICS-KFUPM 27 Learning Agents • • How does an agent improve over time? By monitoring it’s performance and suggesting better modeling, new action rules, Evaluates current world state changes action rules Model world And decide on actions to be taken suggests explorations Dr. Tarek Helmy, ICS-KFUPM 28 Summary What Is Agent?: Anything that can be viewed as perceiving its environment through sensors and acting upon that environment through its effectors to maximize progress towards its goals. Rational Agent Rational Action: The action that maximizes the expected value of the performance measure given the percept sequence to date. Environment Types: Environments are accessible, non-deterministic, non-episodic, dynamic, and continuous. Examples: Agents and Task Environments PEAS (Performance measure, Environment, Actuators, Sensors) Agent Types  Table Driven Agents use a percept-action table in memory to find the next     action. Reflex agents respond immediately to percepts. Goal-based agents act in order to achieve their goals. Utility-based agents maximize their own utility function. Learning agents improve their performance over time. Dr. Tarek Helmy, ICS-KFUPM 29 The End!! Thank you Any Questions? Dr. Tarek Helmy, ICS-KFUPM 30
